The Linux Kernel Mailing List
 help / color / mirror / Atom feed
* linux-next: manual merge of the drm-xe tree with the drm-fixes tree
@ 2024-12-06 12:55 Mark Brown
  0 siblings, 0 replies; 12+ messages in thread
From: Mark Brown @ 2024-12-06 12:55 UTC (permalink / raw)
  To: Lucas De Marchi, Thomas Hellström, DRM XE List
  Cc: John Harrison, Linux Kernel Mailing List, Linux Next Mailing List

[-- Attachment #1: Type: text/plain, Size: 992 bytes --]

Hi all,

Today's linux-next merge of the drm-xe tree got a conflict in:

  drivers/gpu/drm/xe/xe_devcoredump.c

between commit:

  5dce85fecb877 ("drm/xe: Move the coredump registration to the worker thread")

from the drm-fixes tree and commits:

  90f51a7f4ec10 ("drm/xe: Move the coredump registration to the worker thread")
  906c4b306e934 ("drm/xe: Add mutex locking to devcoredump")

from the drm-xe tree.

I fixed it up (see below) and can carry the fix as necessary. This
is now fixed as far as linux-next is concerned, but any non trivial
conflicts should be mentioned to your upstream maintainer when your tree
is submitted for merging.  You may also want to consider cooperating
with the maintainer of the conflicting tree to minimise any particularly
complex conflicts.


diff --cc drivers/gpu/drm/xe/xe_devcoredump.c
index f8947e7e917ec,71636e80b71da..0000000000000
--- a/drivers/gpu/drm/xe/xe_devcoredump.c
+++ b/drivers/gpu/drm/xe/xe_devcoredump.c

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

^ permalink raw reply	[flat|nested] 12+ messages in thread

* linux-next: manual merge of the drm-xe tree with the drm-fixes tree
@ 2025-09-19 12:45 Mark Brown
  2025-09-22 18:28 ` Nathan Chancellor
  0 siblings, 1 reply; 12+ messages in thread
From: Mark Brown @ 2025-09-19 12:45 UTC (permalink / raw)
  To: Lucas De Marchi, Thomas Hellström, DRM XE List
  Cc: Linux Kernel Mailing List, Linux Next Mailing List,
	Michal Wajdeczko, Rodrigo Vivi, Zongyao Bai

[-- Attachment #1: Type: text/plain, Size: 1632 bytes --]

Hi all,

Today's linux-next merge of the drm-xe tree got a conflict in:

  drivers/gpu/drm/xe/xe_guc_submit.c

between commit:

  ff89a4d285c82 ("drm/xe/sysfs: Add cleanup action in xe_device_sysfs_init")

from the drm-fixes tree and commit:

  fb3c27a69c473 ("drm/xe/sysfs: Simplify sysfs registration")

from the drm-xe tree.

I fixed it up (see below) and can carry the fix as necessary. This
is now fixed as far as linux-next is concerned, but any non trivial
conflicts should be mentioned to your upstream maintainer when your tree
is submitted for merging.  You may also want to consider cooperating
with the maintainer of the conflicting tree to minimise any particularly
complex conflicts.

diff --cc drivers/gpu/drm/xe/xe_device_sysfs.c
index b7f8fcfed8d86,c5151c86a98ae..0000000000000
--- a/drivers/gpu/drm/xe/xe_device_sysfs.c
+++ b/drivers/gpu/drm/xe/xe_device_sysfs.c
@@@ -308,19 -290,15 +290,15 @@@ int xe_device_sysfs_init(struct xe_devi
  			return ret;
  	}
  
- 	if (xe->info.platform == XE_BATTLEMAGE) {
- 		ret = sysfs_create_files(&dev->kobj, auto_link_downgrade_attrs);
+ 	if (xe->info.platform == XE_BATTLEMAGE && !IS_SRIOV_VF(xe)) {
+ 		ret = devm_device_add_group(dev, &auto_link_downgrade_attr_group);
  		if (ret)
 -			return ret;
 +			goto cleanup;
  
- 		ret = late_bind_create_files(dev);
+ 		ret = devm_device_add_group(dev, &late_bind_attr_group);
  		if (ret)
 -			return ret;
 +			goto cleanup;
  	}
  
- 	return devm_add_action_or_reset(dev, xe_device_sysfs_fini, xe);
- 
- cleanup:
- 	xe_device_sysfs_fini(xe);
- 	return ret;
+ 	return 0;
  }

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

^ permalink raw reply	[flat|nested] 12+ messages in thread

* linux-next: manual merge of the drm-xe tree with the drm-fixes tree
@ 2025-09-19 12:49 Mark Brown
  0 siblings, 0 replies; 12+ messages in thread
From: Mark Brown @ 2025-09-19 12:49 UTC (permalink / raw)
  To: Lucas De Marchi, Thomas Hellström, DRM XE List
  Cc: Daniele Ceraolo Spurio, Linux Kernel Mailing List,
	Linux Next Mailing List, Rodrigo Vivi

[-- Attachment #1: Type: text/plain, Size: 877 bytes --]

Hi all,

Today's linux-next merge of the drm-xe tree got a conflict in:

  drivers/gpu/drm/xe/xe_guc_submit.c

between commit:

  26caeae9fb482 ("drm/xe/guc: Set RCS/CCS yield policy")

from the drm-fixes tree and commit:

  88434448438e4 ("drm/xe/guc: Set RCS/CCS yield policy")

from the drm-xe tree.

I fixed it up (see below) and can carry the fix as necessary. This
is now fixed as far as linux-next is concerned, but any non trivial
conflicts should be mentioned to your upstream maintainer when your tree
is submitted for merging.  You may also want to consider cooperating
with the maintainer of the conflicting tree to minimise any particularly
complex conflicts.

diff --cc drivers/gpu/drm/xe/xe_guc_submit.c
index e377ba3a39b3a,53024eb5670b7..0000000000000
--- a/drivers/gpu/drm/xe/xe_guc_submit.c
+++ b/drivers/gpu/drm/xe/xe_guc_submit.c

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

^ permalink raw reply	[flat|nested] 12+ messages in thread

* Re: linux-next: manual merge of the drm-xe tree with the drm-fixes tree
  2025-09-19 12:45 Mark Brown
@ 2025-09-22 18:28 ` Nathan Chancellor
  2025-09-22 19:25   ` Lucas De Marchi
  0 siblings, 1 reply; 12+ messages in thread
From: Nathan Chancellor @ 2025-09-22 18:28 UTC (permalink / raw)
  To: Mark Brown
  Cc: Lucas De Marchi, Thomas Hellström, DRM XE List,
	Linux Kernel Mailing List, Linux Next Mailing List,
	Michal Wajdeczko, Rodrigo Vivi, Zongyao Bai

Hi Mark,

On Fri, Sep 19, 2025 at 01:45:19PM +0100, Mark Brown wrote:
> Hi all,
> 
> Today's linux-next merge of the drm-xe tree got a conflict in:
> 
>   drivers/gpu/drm/xe/xe_guc_submit.c
> 
> between commit:
> 
>   ff89a4d285c82 ("drm/xe/sysfs: Add cleanup action in xe_device_sysfs_init")
> 
> from the drm-fixes tree and commit:
> 
>   fb3c27a69c473 ("drm/xe/sysfs: Simplify sysfs registration")
> 
> from the drm-xe tree.
> 
> I fixed it up (see below) and can carry the fix as necessary. This
> is now fixed as far as linux-next is concerned, but any non trivial
> conflicts should be mentioned to your upstream maintainer when your tree
> is submitted for merging.  You may also want to consider cooperating
> with the maintainer of the conflicting tree to minimise any particularly
> complex conflicts.
> 
> diff --cc drivers/gpu/drm/xe/xe_device_sysfs.c
> index b7f8fcfed8d86,c5151c86a98ae..0000000000000
> --- a/drivers/gpu/drm/xe/xe_device_sysfs.c
> +++ b/drivers/gpu/drm/xe/xe_device_sysfs.c
> @@@ -308,19 -290,15 +290,15 @@@ int xe_device_sysfs_init(struct xe_devi
>   			return ret;
>   	}
>   
> - 	if (xe->info.platform == XE_BATTLEMAGE) {
> - 		ret = sysfs_create_files(&dev->kobj, auto_link_downgrade_attrs);
> + 	if (xe->info.platform == XE_BATTLEMAGE && !IS_SRIOV_VF(xe)) {
> + 		ret = devm_device_add_group(dev, &auto_link_downgrade_attr_group);
>   		if (ret)
>  -			return ret;
>  +			goto cleanup;
>   
> - 		ret = late_bind_create_files(dev);
> + 		ret = devm_device_add_group(dev, &late_bind_attr_group);
>   		if (ret)
>  -			return ret;
>  +			goto cleanup;
>   	}
>   
> - 	return devm_add_action_or_reset(dev, xe_device_sysfs_fini, xe);
> - 
> - cleanup:
> - 	xe_device_sysfs_fini(xe);
> - 	return ret;
> + 	return 0;
>   }

This resolution breaks the build.

  drivers/gpu/drm/xe/xe_device_sysfs.c: In function 'xe_device_sysfs_init':
  drivers/gpu/drm/xe/xe_device_sysfs.c:300:25: error: label 'cleanup' used but not defined
    300 |                         goto cleanup;
        |                         ^~~~

I think those should just be 'return ret', as it appears that
fb3c27a69c473 already includes ff89a4d285c82, so I think you can just
take the right side of the conflict wholesale.

Cheers,
Nathan

^ permalink raw reply	[flat|nested] 12+ messages in thread

* Re: linux-next: manual merge of the drm-xe tree with the drm-fixes tree
  2025-09-22 18:28 ` Nathan Chancellor
@ 2025-09-22 19:25   ` Lucas De Marchi
  0 siblings, 0 replies; 12+ messages in thread
From: Lucas De Marchi @ 2025-09-22 19:25 UTC (permalink / raw)
  To: Nathan Chancellor
  Cc: Mark Brown, Thomas Hellström, DRM XE List,
	Linux Kernel Mailing List, Linux Next Mailing List,
	Michal Wajdeczko, Rodrigo Vivi, Zongyao Bai

On Mon, Sep 22, 2025 at 11:28:32AM -0700, Nathan Chancellor wrote:
>Hi Mark,
>
>On Fri, Sep 19, 2025 at 01:45:19PM +0100, Mark Brown wrote:
>> Hi all,
>>
>> Today's linux-next merge of the drm-xe tree got a conflict in:
>>
>>   drivers/gpu/drm/xe/xe_guc_submit.c
>>
>> between commit:
>>
>>   ff89a4d285c82 ("drm/xe/sysfs: Add cleanup action in xe_device_sysfs_init")
>>
>> from the drm-fixes tree and commit:
>>
>>   fb3c27a69c473 ("drm/xe/sysfs: Simplify sysfs registration")
>>
>> from the drm-xe tree.
>>
>> I fixed it up (see below) and can carry the fix as necessary. This
>> is now fixed as far as linux-next is concerned, but any non trivial
>> conflicts should be mentioned to your upstream maintainer when your tree
>> is submitted for merging.  You may also want to consider cooperating
>> with the maintainer of the conflicting tree to minimise any particularly
>> complex conflicts.
>>
>> diff --cc drivers/gpu/drm/xe/xe_device_sysfs.c
>> index b7f8fcfed8d86,c5151c86a98ae..0000000000000
>> --- a/drivers/gpu/drm/xe/xe_device_sysfs.c
>> +++ b/drivers/gpu/drm/xe/xe_device_sysfs.c
>> @@@ -308,19 -290,15 +290,15 @@@ int xe_device_sysfs_init(struct xe_devi
>>   			return ret;
>>   	}
>>
>> - 	if (xe->info.platform == XE_BATTLEMAGE) {
>> - 		ret = sysfs_create_files(&dev->kobj, auto_link_downgrade_attrs);
>> + 	if (xe->info.platform == XE_BATTLEMAGE && !IS_SRIOV_VF(xe)) {
>> + 		ret = devm_device_add_group(dev, &auto_link_downgrade_attr_group);
>>   		if (ret)
>>  -			return ret;
>>  +			goto cleanup;
>>
>> - 		ret = late_bind_create_files(dev);
>> + 		ret = devm_device_add_group(dev, &late_bind_attr_group);
>>   		if (ret)
>>  -			return ret;
>>  +			goto cleanup;
>>   	}
>>
>> - 	return devm_add_action_or_reset(dev, xe_device_sysfs_fini, xe);
>> -
>> - cleanup:
>> - 	xe_device_sysfs_fini(xe);
>> - 	return ret;
>> + 	return 0;
>>   }
>
>This resolution breaks the build.
>
>  drivers/gpu/drm/xe/xe_device_sysfs.c: In function 'xe_device_sysfs_init':
>  drivers/gpu/drm/xe/xe_device_sysfs.c:300:25: error: label 'cleanup' used but not defined
>    300 |                         goto cleanup;
>        |                         ^~~~
>
>I think those should just be 'return ret', as it appears that
>fb3c27a69c473 already includes ff89a4d285c82, so I think you can just
>take the right side of the conflict wholesale.

yes, taking the right side is the correct resolution.

thanks
Lucas De Marchi

>
>Cheers,
>Nathan

^ permalink raw reply	[flat|nested] 12+ messages in thread

* linux-next: manual merge of the drm-xe tree with the drm-fixes tree
@ 2026-01-23 15:34 Mark Brown
  0 siblings, 0 replies; 12+ messages in thread
From: Mark Brown @ 2026-01-23 15:34 UTC (permalink / raw)
  To: Lucas De Marchi, Thomas Hellström, DRM XE List
  Cc: Linux Kernel Mailing List, Linux Next Mailing List

Hi all,

Today's linux-next merge of the drm-xe tree got a conflict in:

  include/drm/drm_pagemap.h

between commit:

  bdcdf968be314 ("drm, drm/xe: Fix xe userptr in the absence of CONFIG_DEVICE_PRIVATE")

from the drm-fixes tree and commit:

  1e372b246199c ("drm, drm/xe: Fix xe userptr in the absence of CONFIG_DEVICE_PRIVATE")

from the drm-xe tree.

I fixed it up (see below) and can carry the fix as necessary. This
is now fixed as far as linux-next is concerned, but any non trivial
conflicts should be mentioned to your upstream maintainer when your tree
is submitted for merging.  You may also want to consider cooperating
with the maintainer of the conflicting tree to minimise any particularly
complex conflicts.

diff --cc include/drm/drm_pagemap.h
index b7a78c1ca7670,2baf0861f78f3..0000000000000
--- a/include/drm/drm_pagemap.h
+++ b/include/drm/drm_pagemap.h

^ permalink raw reply	[flat|nested] 12+ messages in thread

* linux-next: manual merge of the drm-xe tree with the drm-fixes tree
@ 2026-01-30 16:05 Mark Brown
  0 siblings, 0 replies; 12+ messages in thread
From: Mark Brown @ 2026-01-30 16:05 UTC (permalink / raw)
  To: Thomas Hellström, DRM XE List
  Cc: Ashutosh Dixit, Linux Kernel Mailing List,
	Linux Next Mailing List, Shuicheng Lin

[-- Attachment #1: Type: text/plain, Size: 1927 bytes --]

Hi all,

Today's linux-next merge of the drm-xe tree got a conflict in:

  drivers/gpu/drm/xe/xe_nvm.c

between commit:

  8a44241b0b83a ("drm/xe/nvm: Fix double-free on aux add failure")

from the drm-fixes tree and commits:

  11035eab1b7d8 ("drm/xe/nvm: Manage nvm aux cleanup with devres")
  a3187c0c2bbd9 ("drm/xe/nvm: Fix double-free on aux add failure")
  7755ed58a49f4 ("drm/xe/nvm: Defer xe->nvm assignment until init succeeds")

from the drm-xe tree.

I fixed it up (see below) and can carry the fix as necessary. This
is now fixed as far as linux-next is concerned, but any non trivial
conflicts should be mentioned to your upstream maintainer when your tree
is submitted for merging.  You may also want to consider cooperating
with the maintainer of the conflicting tree to minimise any particularly
complex conflicts.

diff --combined drivers/gpu/drm/xe/xe_nvm.c
index 6f9dd519371c9,bc88804de5146..0000000000000
--- a/drivers/gpu/drm/xe/xe_nvm.c
+++ b/drivers/gpu/drm/xe/xe_nvm.c
@@@ -133,12 -133,10 +133,10 @@@ int xe_nvm_init(struct xe_device *xe
  	if (WARN_ON(xe->nvm))
  		return -EFAULT;
  
- 	xe->nvm = kzalloc(sizeof(*nvm), GFP_KERNEL);
- 	if (!xe->nvm)
+ 	nvm = kzalloc(sizeof(*nvm), GFP_KERNEL);
+ 	if (!nvm)
  		return -ENOMEM;
  
- 	nvm = xe->nvm;
- 
  	nvm->writable_override = xe_nvm_writable_override(xe);
  	nvm->non_posted_erase = xe_nvm_non_posted_erase(xe);
  	nvm->bar.parent = &pdev->resource[0];
@@@ -165,7 -163,6 +163,6 @@@
  	if (ret) {
  		drm_err(&xe->drm, "xe-nvm aux init failed %d\n", ret);
  		kfree(nvm);
- 		xe->nvm = NULL;
  		return ret;
  	}
  
@@@ -173,8 -170,9 +170,9 @@@
  	if (ret) {
  		drm_err(&xe->drm, "xe-nvm aux add failed %d\n", ret);
  		auxiliary_device_uninit(aux_dev);
- 		xe->nvm = NULL;
  		return ret;
  	}
+ 
+ 	xe->nvm = nvm;
  	return devm_add_action_or_reset(xe->drm.dev, xe_nvm_fini, xe);
  }

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

^ permalink raw reply	[flat|nested] 12+ messages in thread

* linux-next: manual merge of the drm-xe tree with the drm-fixes tree
@ 2026-07-03 15:41 Mark Brown
  0 siblings, 0 replies; 12+ messages in thread
From: Mark Brown @ 2026-07-03 15:41 UTC (permalink / raw)
  To: Thomas Hellström, DRM XE List
  Cc: Gustavo Sousa, Linux Kernel Mailing List, Linux Next Mailing List,
	Matt Roper, Violet Monti

[-- Attachment #1: Type: text/plain, Size: 10917 bytes --]

Hi all,

Today's linux-next merge of the drm-xe tree got a conflict in:

  drivers/gpu/drm/xe/xe_hw_engine.c

between commit:

  e23fafb8594ea ("drm/xe/rtp: Add struct types for RTP tables")

from the drm-fixes tree and commits:

  5ff004fdc7377 ("drm/xe/rtp: Add struct types for RTP tables")
  431a233c1710c ("drm/xe: Move engines' LRC programming RTP table off the stack")
  4ff7902a64c18 ("drm/xe: Move engines' non-LRC programming RTP table off the stack")

from the drm-xe tree.

I fixed it up (see below) and can carry the fix as necessary. This
is now fixed as far as linux-next is concerned, but any non trivial
conflicts should be mentioned to your upstream maintainer when your tree
is submitted for merging.  You may also want to consider cooperating
with the maintainer of the conflicting tree to minimise any particularly
complex conflicts.

diff --combined drivers/gpu/drm/xe/xe_hw_engine.c
index 0b193c451a11d,87d60c4117bd8..0000000000000
--- a/drivers/gpu/drm/xe/xe_hw_engine.c
+++ b/drivers/gpu/drm/xe/xe_hw_engine.c
@@@ -337,39 -337,41 +337,41 @@@ static bool xe_rtp_cfeg_wmtp_disabled(c
  	return xe_mmio_read32(&hwe->gt->mmio, XEHP_FUSE4) & CFEG_WMTP_DISABLE;
  }
  
+ static u32 blit_cctl_val(struct xe_gt *gt, struct xe_hw_engine *hwe)
+ {
+ 	return REG_FIELD_PREP(BLIT_CCTL_DST_MOCS_MASK, gt->mocs.uc_index) |
+ 		REG_FIELD_PREP(BLIT_CCTL_SRC_MOCS_MASK, gt->mocs.uc_index);
+ }
+ 
+ static const struct xe_rtp_table_sr lrc_setup = XE_RTP_TABLE_SR(
+ 	/*
+ 	 * Some blitter commands do not have a field for MOCS, those
+ 	 * commands will use MOCS index pointed by BLIT_CCTL.
+ 	 * BLIT_CCTL registers are needed to be programmed to un-cached.
+ 	 */
+ 	{ XE_RTP_NAME("BLIT_CCTL_default_MOCS"),
+ 	  XE_RTP_RULES(GRAPHICS_VERSION_RANGE(1200, 1274),
+ 		       ENGINE_CLASS(COPY)),
+ 	  XE_RTP_ACTIONS(FIELD_SET_FUNC(BLIT_CCTL(0),
+ 					BLIT_CCTL_DST_MOCS_MASK |
+ 					BLIT_CCTL_SRC_MOCS_MASK,
+ 					blit_cctl_val,
+ 					XE_RTP_ACTION_FLAG(ENGINE_BASE)))
+ 	},
+ 	/* Disable WMTP if HW doesn't support it */
+ 	{ XE_RTP_NAME("DISABLE_WMTP_ON_UNSUPPORTED_HW"),
+ 	  XE_RTP_RULES(FUNC(xe_rtp_cfeg_wmtp_disabled)),
+ 	  XE_RTP_ACTIONS(FIELD_SET(CS_CHICKEN1(0),
+ 				   PREEMPT_GPGPU_LEVEL_MASK,
+ 				   PREEMPT_GPGPU_THREAD_GROUP_LEVEL)),
+ 	  XE_RTP_ENTRY_FLAG(FOREACH_ENGINE)
+ 	},
+ );
+ 
  static void
  hw_engine_setup_default_lrc_state(struct xe_hw_engine *hwe)
  {
- 	struct xe_gt *gt = hwe->gt;
- 	const u8 mocs_write_idx = gt->mocs.uc_index;
- 	const u8 mocs_read_idx = gt->mocs.uc_index;
- 	u32 blit_cctl_val = REG_FIELD_PREP(BLIT_CCTL_DST_MOCS_MASK, mocs_write_idx) |
- 			    REG_FIELD_PREP(BLIT_CCTL_SRC_MOCS_MASK, mocs_read_idx);
  	struct xe_rtp_process_ctx ctx = XE_RTP_PROCESS_CTX_INITIALIZER(hwe);
- 	const struct xe_rtp_table_sr lrc_setup = XE_RTP_TABLE_SR(
- 		/*
- 		 * Some blitter commands do not have a field for MOCS, those
- 		 * commands will use MOCS index pointed by BLIT_CCTL.
- 		 * BLIT_CCTL registers are needed to be programmed to un-cached.
- 		 */
- 		{ XE_RTP_NAME("BLIT_CCTL_default_MOCS"),
- 		  XE_RTP_RULES(GRAPHICS_VERSION_RANGE(1200, 1274),
- 			       ENGINE_CLASS(COPY)),
- 		  XE_RTP_ACTIONS(FIELD_SET(BLIT_CCTL(0),
- 				 BLIT_CCTL_DST_MOCS_MASK |
- 				 BLIT_CCTL_SRC_MOCS_MASK,
- 				 blit_cctl_val,
- 				 XE_RTP_ACTION_FLAG(ENGINE_BASE)))
- 		},
- 		/* Disable WMTP if HW doesn't support it */
- 		{ XE_RTP_NAME("DISABLE_WMTP_ON_UNSUPPORTED_HW"),
- 		  XE_RTP_RULES(FUNC(xe_rtp_cfeg_wmtp_disabled)),
- 		  XE_RTP_ACTIONS(FIELD_SET(CS_CHICKEN1(0),
- 					   PREEMPT_GPGPU_LEVEL_MASK,
- 					   PREEMPT_GPGPU_THREAD_GROUP_LEVEL)),
- 		  XE_RTP_ENTRY_FLAG(FOREACH_ENGINE)
- 		},
- 	);
  
  	xe_rtp_process_to_sr(&ctx, &lrc_setup, &hwe->reg_lrc, true);
  }
@@@ -385,86 -387,92 +387,92 @@@ void xe_hw_engine_setup_reg_lrc(struct 
  	xe_tuning_process_lrc(hwe);
  }
  
+ /*
+  * RING_CMD_CCTL specifies the default MOCS entry that will be
+  * used by the command streamer when executing commands that
+  * don't have a way to explicitly specify a MOCS setting.
+  * The default should usually reference whichever MOCS entry
+  * corresponds to uncached behavior, although use of a WB cached
+  * entry is recommended by the spec in certain circumstances on
+  * specific platforms.
+  * Bspec: 72161
+  */
+ static u32 ring_cmd_cctl_val(struct xe_gt *gt, struct xe_hw_engine *hwe)
+ {
+ 	struct xe_device *xe = gt_to_xe(gt);
+ 	u8 mocs_read_idx = gt->mocs.uc_index;
+ 
+ 	if (hwe->class == XE_ENGINE_CLASS_COMPUTE && IS_DGFX(xe) &&
+ 	    (GRAPHICS_VER(xe) >= 20 || xe->info.platform == XE_PVC))
+ 		mocs_read_idx = gt->mocs.wb_index;
+ 
+ 	return REG_FIELD_PREP(CMD_CCTL_WRITE_OVERRIDE_MASK, gt->mocs.uc_index) |
+ 		REG_FIELD_PREP(CMD_CCTL_READ_OVERRIDE_MASK, mocs_read_idx);
+ }
+ 
+ static const struct xe_rtp_table_sr engine_sr = XE_RTP_TABLE_SR(
+ 	{ XE_RTP_NAME("RING_CMD_CCTL_default_MOCS"),
+ 	  XE_RTP_RULES(FUNC(xe_rtp_match_always)),
+ 	  XE_RTP_ACTIONS(FIELD_SET_FUNC(RING_CMD_CCTL(0),
+ 					CMD_CCTL_WRITE_OVERRIDE_MASK |
+ 					CMD_CCTL_READ_OVERRIDE_MASK,
+ 					ring_cmd_cctl_val,
+ 					XE_RTP_ACTION_FLAG(ENGINE_BASE)))
+ 	},
+ 	{ XE_RTP_NAME("Disable HW status page updates for interrupts"),
+ 	  XE_RTP_RULES(FUNC(xe_rtp_match_always)),
+ 	  XE_RTP_ACTIONS(SET(RING_HWSTAM(0), ~0x0,
+ 			     XE_RTP_ACTION_FLAG(ENGINE_BASE)))
+ 	},
+ 	{ XE_RTP_NAME("Disable engine 'legacy' mode"),
+ 	  XE_RTP_RULES(FUNC(xe_rtp_match_always)),
+ 	  XE_RTP_ACTIONS(SET(GFX_MODE(0), GFX_DISABLE_LEGACY_MODE,
+ 			     XE_RTP_ACTION_FLAG(ENGINE_BASE)))
+ 	},
+ 	/*
+ 	 * To allow the GSC engine to go idle on MTL we need to enable
+ 	 * idle messaging and set the hysteresis value (we use 0xA=5us
+ 	 * as recommended in spec). On platforms after MTL this is
+ 	 * enabled by default.
+ 	 */
+ 	{ XE_RTP_NAME("MTL GSCCS IDLE MSG enable"),
+ 	  XE_RTP_RULES(MEDIA_VERSION(1300), ENGINE_CLASS(OTHER)),
+ 	  XE_RTP_ACTIONS(CLR(RING_PSMI_CTL(0),
+ 			     IDLE_MSG_DISABLE,
+ 			     XE_RTP_ACTION_FLAG(ENGINE_BASE)),
+ 			 FIELD_SET(RING_PWRCTX_MAXCNT(0),
+ 				   IDLE_WAIT_TIME,
+ 				   0xA,
+ 				   XE_RTP_ACTION_FLAG(ENGINE_BASE)))
+ 	},
+ 	/* Enable Priority Mem Read */
+ 	{ XE_RTP_NAME("Priority_Mem_Read"),
+ 	  XE_RTP_RULES(GRAPHICS_VERSION_RANGE(2001, XE_RTP_END_VERSION_UNDEFINED)),
+ 	  XE_RTP_ACTIONS(SET(CSFE_CHICKEN1(0), CS_PRIORITY_MEM_READ,
+ 			     XE_RTP_ACTION_FLAG(ENGINE_BASE)))
+ 	},
+ 	{ XE_RTP_NAME("Enable CCS Engine(s)"),
+ 	  XE_RTP_RULES(GRAPHICS_VERSION_RANGE(1255, XE_RTP_END_VERSION_UNDEFINED),
+ 		       FUNC(xe_rtp_match_first_render_or_compute)),
+ 	  XE_RTP_ACTIONS(SET(RCU_MODE, RCU_MODE_CCS_ENABLE))
+ 	},
+ 	/* Use Fixed slice CCS mode */
+ 	{ XE_RTP_NAME("RCU_MODE_FIXED_SLICE_CCS_MODE"),
+ 	  XE_RTP_RULES(FUNC(xe_hw_engine_match_fixed_cslice_mode)),
+ 	  XE_RTP_ACTIONS(FIELD_SET(RCU_MODE, RCU_MODE_FIXED_SLICE_CCS_MODE,
+ 				   RCU_MODE_FIXED_SLICE_CCS_MODE))
+ 	},
+ 	{ XE_RTP_NAME("Enable MSI-X interrupt support"),
+ 	  XE_RTP_RULES(FUNC(xe_rtp_match_has_msix)),
+ 	  XE_RTP_ACTIONS(SET(GFX_MODE(0), GFX_MSIX_INTERRUPT_ENABLE,
+ 			     XE_RTP_ACTION_FLAG(ENGINE_BASE)))
+ 	},
+ );
+ 
  static void
  hw_engine_setup_default_state(struct xe_hw_engine *hwe)
  {
- 	struct xe_gt *gt = hwe->gt;
- 	struct xe_device *xe = gt_to_xe(gt);
- 	/*
- 	 * RING_CMD_CCTL specifies the default MOCS entry that will be
- 	 * used by the command streamer when executing commands that
- 	 * don't have a way to explicitly specify a MOCS setting.
- 	 * The default should usually reference whichever MOCS entry
- 	 * corresponds to uncached behavior, although use of a WB cached
- 	 * entry is recommended by the spec in certain circumstances on
- 	 * specific platforms.
- 	 * Bspec: 72161
- 	 */
- 	const u8 mocs_write_idx = gt->mocs.uc_index;
- 	const u8 mocs_read_idx = hwe->class == XE_ENGINE_CLASS_COMPUTE && IS_DGFX(xe) &&
- 				 (GRAPHICS_VER(xe) >= 20 || xe->info.platform == XE_PVC) ?
- 				 gt->mocs.wb_index : gt->mocs.uc_index;
- 	u32 ring_cmd_cctl_val = REG_FIELD_PREP(CMD_CCTL_WRITE_OVERRIDE_MASK, mocs_write_idx) |
- 				REG_FIELD_PREP(CMD_CCTL_READ_OVERRIDE_MASK, mocs_read_idx);
  	struct xe_rtp_process_ctx ctx = XE_RTP_PROCESS_CTX_INITIALIZER(hwe);
- 	const struct xe_rtp_table_sr engine_sr = XE_RTP_TABLE_SR(
- 		{ XE_RTP_NAME("RING_CMD_CCTL_default_MOCS"),
- 		  XE_RTP_RULES(FUNC(xe_rtp_match_always)),
- 		  XE_RTP_ACTIONS(FIELD_SET(RING_CMD_CCTL(0),
- 					   CMD_CCTL_WRITE_OVERRIDE_MASK |
- 					   CMD_CCTL_READ_OVERRIDE_MASK,
- 					   ring_cmd_cctl_val,
- 					   XE_RTP_ACTION_FLAG(ENGINE_BASE)))
- 		},
- 		{ XE_RTP_NAME("Disable HW status page updates for interrupts"),
- 		  XE_RTP_RULES(FUNC(xe_rtp_match_always)),
- 		  XE_RTP_ACTIONS(SET(RING_HWSTAM(0), ~0x0,
- 				     XE_RTP_ACTION_FLAG(ENGINE_BASE)))
- 		},
- 		{ XE_RTP_NAME("Disable engine 'legacy' mode"),
- 		  XE_RTP_RULES(FUNC(xe_rtp_match_always)),
- 		  XE_RTP_ACTIONS(SET(GFX_MODE(0), GFX_DISABLE_LEGACY_MODE,
- 				     XE_RTP_ACTION_FLAG(ENGINE_BASE)))
- 		},
- 		/*
- 		 * To allow the GSC engine to go idle on MTL we need to enable
- 		 * idle messaging and set the hysteresis value (we use 0xA=5us
- 		 * as recommended in spec). On platforms after MTL this is
- 		 * enabled by default.
- 		 */
- 		{ XE_RTP_NAME("MTL GSCCS IDLE MSG enable"),
- 		  XE_RTP_RULES(MEDIA_VERSION(1300), ENGINE_CLASS(OTHER)),
- 		  XE_RTP_ACTIONS(CLR(RING_PSMI_CTL(0),
- 				     IDLE_MSG_DISABLE,
- 				     XE_RTP_ACTION_FLAG(ENGINE_BASE)),
- 				 FIELD_SET(RING_PWRCTX_MAXCNT(0),
- 					   IDLE_WAIT_TIME,
- 					   0xA,
- 					   XE_RTP_ACTION_FLAG(ENGINE_BASE)))
- 		},
- 		/* Enable Priority Mem Read */
- 		{ XE_RTP_NAME("Priority_Mem_Read"),
- 		  XE_RTP_RULES(GRAPHICS_VERSION_RANGE(2001, XE_RTP_END_VERSION_UNDEFINED)),
- 		  XE_RTP_ACTIONS(SET(CSFE_CHICKEN1(0), CS_PRIORITY_MEM_READ,
- 				     XE_RTP_ACTION_FLAG(ENGINE_BASE)))
- 		},
- 		{ XE_RTP_NAME("Enable CCS Engine(s)"),
- 		  XE_RTP_RULES(GRAPHICS_VERSION_RANGE(1255, XE_RTP_END_VERSION_UNDEFINED),
- 			       FUNC(xe_rtp_match_first_render_or_compute)),
- 		  XE_RTP_ACTIONS(SET(RCU_MODE, RCU_MODE_CCS_ENABLE))
- 		},
- 		/* Use Fixed slice CCS mode */
- 		{ XE_RTP_NAME("RCU_MODE_FIXED_SLICE_CCS_MODE"),
- 		  XE_RTP_RULES(FUNC(xe_hw_engine_match_fixed_cslice_mode)),
- 		  XE_RTP_ACTIONS(FIELD_SET(RCU_MODE, RCU_MODE_FIXED_SLICE_CCS_MODE,
- 					   RCU_MODE_FIXED_SLICE_CCS_MODE))
- 		},
- 		{ XE_RTP_NAME("Enable MSI-X interrupt support"),
- 		  XE_RTP_RULES(FUNC(xe_rtp_match_has_msix)),
- 		  XE_RTP_ACTIONS(SET(GFX_MODE(0), GFX_MSIX_INTERRUPT_ENABLE,
- 				     XE_RTP_ACTION_FLAG(ENGINE_BASE)))
- 		},
- 	);
  
  	xe_rtp_process_to_sr(&ctx, &engine_sr, &hwe->reg_sr, false);
  }

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

^ permalink raw reply	[flat|nested] 12+ messages in thread

* linux-next: manual merge of the drm-xe tree with the drm-fixes tree
@ 2026-07-03 15:41 Mark Brown
  0 siblings, 0 replies; 12+ messages in thread
From: Mark Brown @ 2026-07-03 15:41 UTC (permalink / raw)
  To: Thomas Hellström, DRM XE List
  Cc: Gustavo Sousa, Linux Kernel Mailing List, Linux Next Mailing List,
	Matt Roper, Violet Monti

[-- Attachment #1: Type: text/plain, Size: 3259 bytes --]

Hi all,

Today's linux-next merge of the drm-xe tree got a conflict in:

  drivers/gpu/drm/xe/tests/xe_rtp_test.c

between commit:

  e23fafb8594ea ("drm/xe/rtp: Add struct types for RTP tables")

from the drm-fixes tree and commits:

  5ff004fdc7377 ("drm/xe/rtp: Add struct types for RTP tables")
  c47ffed42b016 ("drm/xe/rtp: Add kunit tests to exercise FIELD_SET_FUNC action")

from the drm-xe tree.

I fixed it up (see below) and can carry the fix as necessary. This
is now fixed as far as linux-next is concerned, but any non trivial
conflicts should be mentioned to your upstream maintainer when your tree
is submitted for merging.  You may also want to consider cooperating
with the maintainer of the conflicting tree to minimise any particularly
complex conflicts.

diff --combined drivers/gpu/drm/xe/tests/xe_rtp_test.c
index 3d0688d058d97,367811621880f..0000000000000
--- a/drivers/gpu/drm/xe/tests/xe_rtp_test.c
+++ b/drivers/gpu/drm/xe/tests/xe_rtp_test.c
@@@ -280,6 -280,11 +280,11 @@@ static void xe_rtp_rules_tests(struct k
  	KUNIT_EXPECT_EQ(test, err, param->expected_err);
  }
  
+ static u32 bits_2_3_set(struct xe_gt *gt, struct xe_hw_engine *hwe)
+ {
+ 	return REG_BIT(2) | REG_BIT(3);
+ }
+ 
  static const struct rtp_to_sr_test_case rtp_to_sr_cases[] = {
  	{
  		.name = "coalesce-same-reg",
@@@ -300,6 -305,29 +305,29 @@@
  			},
  		),
  	},
+ 	{
+ 		.name = "coalesce-same-reg-literal-and-func",
+ 		.expected_reg = REGULAR_REG1,
+ 		.expected_set_bits = REG_BIT(0) | REG_BIT(1) | REG_BIT(2) | REG_BIT(3),
+ 		.expected_clr_bits = REG_BIT(0) | REG_BIT(1) | REG_BIT(2) | REG_BIT(3),
+ 		.expected_active = BIT(0) | BIT(1),
+ 		.expected_count_sr_entries = 1,
+ 		/* Different bits on the same register: create a single entry */
+ 		.table = XE_RTP_TABLE_SR(
+ 			{ XE_RTP_NAME("basic-1"),
+ 			  XE_RTP_RULES(FUNC(match_yes)),
+ 			  XE_RTP_ACTIONS(FIELD_SET(REGULAR_REG1,
+ 						   REG_BIT(0) | REG_BIT(1),
+ 						   REG_BIT(0) | REG_BIT(1)))
+ 			},
+ 			{ XE_RTP_NAME("basic-2"),
+ 			  XE_RTP_RULES(FUNC(match_yes)),
+ 			  XE_RTP_ACTIONS(FIELD_SET_FUNC(REGULAR_REG1,
+ 							REG_BIT(2) | REG_BIT(3),
+ 							bits_2_3_set))
+ 			},
+ 		),
+ 	},
  	{
  		.name = "no-match-no-add",
  		.expected_reg = REGULAR_REG1,
@@@ -417,6 -445,30 +445,30 @@@
  			},
  		),
  	},
+ 	{
+ 		.name = "conflict-not-disjoint-literal-and-func",
+ 		.expected_reg = REGULAR_REG1,
+ 		.expected_set_bits = REG_BIT(1) | REG_BIT(2),
+ 		.expected_clr_bits = REG_BIT(1) | REG_BIT(2),
+ 		.expected_active = BIT(0) | BIT(1),
+ 		.expected_count_sr_entries = 1,
+ 		.expected_sr_errors = 1,
+ 		.table = XE_RTP_TABLE_SR(
+ 			{ XE_RTP_NAME("basic-1"),
+ 			  XE_RTP_RULES(FUNC(match_yes)),
+ 			  XE_RTP_ACTIONS(FIELD_SET(REGULAR_REG1,
+ 						   REG_BIT(1) | REG_BIT(2),
+ 						   REG_BIT(1) | REG_BIT(2)))
+ 			},
+ 			/* drop: bits are not disjoint with previous entries */
+ 			{ XE_RTP_NAME("basic-2"),
+ 			  XE_RTP_RULES(FUNC(match_yes)),
+ 			  XE_RTP_ACTIONS(FIELD_SET_FUNC(REGULAR_REG1,
+ 							REG_BIT(2) | REG_BIT(3),
+ 							bits_2_3_set))
+ 			},
+ 		),
+ 	},
  	{
  		.name = "conflict-reg-type",
  		.expected_reg = REGULAR_REG1,

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

^ permalink raw reply	[flat|nested] 12+ messages in thread

* linux-next: manual merge of the drm-xe tree with the drm-fixes tree
@ 2026-07-03 15:44 Mark Brown
  0 siblings, 0 replies; 12+ messages in thread
From: Mark Brown @ 2026-07-03 15:44 UTC (permalink / raw)
  To: Thomas Hellström, DRM XE List
  Cc: Ashutosh Dixit, Gustavo Sousa, Linux Kernel Mailing List,
	Linux Next Mailing List, Matt Roper, Violet Monti

[-- Attachment #1: Type: text/plain, Size: 2639 bytes --]

Hi all,

Today's linux-next merge of the drm-xe tree got a conflict in:

  drivers/gpu/drm/xe/xe_reg_whitelist.c

between commit:

  959b5016e4646 ("drm/xe/oa: Fix offset alignment for MERT WHITELIST_OA_MERT_MMIO_TRG")

from the drm-fixes tree and commits:

  5ff004fdc7377 ("drm/xe/rtp: Add struct types for RTP tables")
  90511bdcfda97 ("drm/xe/rtp: Add RING_FORCE_TO_NONPRIV_DENY to OA whitelists")
  c478244a9e2d1 ("drm/xe/rtp: Maintain OA whitelists separately")
  15739920b71ef ("drm/xe/rtp: Keep track of non-OA nonpriv slots")
  c3ff77d7235cc ("drm/xe/rtp: Generalize whitelist_apply_to_hwe")
  3a3c3e56db292 ("drm/xe/rtp: Save OA nonpriv registers to register save/restore lists")
  aeaa7d2bb0172 ("drm/xe/rtp: Toggle 'deny' bit to (de-)whitelist OA regs")
  6f73bf8fffa72 ("drm/xe/rtp: (De-)whitelist OA registers for all hwe's for a gt")
  645f1a2589bd4 ("drm/xe/rtp: Ensure locking/ref counting for OA whitelists")
  483c9f5451592 ("drm/xe/tests/rtp: Add kunit test for whitelist upper bounds")
  f6c23e4589bdc ("drm/xe/oa: Fix offset alignment for MERT WHITELIST_OA_MERT_MMIO_TRG")

from the drm-xe tree.

I fixed it up (see below) and can carry the fix as necessary. This
is now fixed as far as linux-next is concerned, but any non trivial
conflicts should be mentioned to your upstream maintainer when your tree
is submitted for merging.  You may also want to consider cooperating
with the maintainer of the conflicting tree to minimise any particularly
complex conflicts.


diff --combined drivers/gpu/drm/xe/xe_reg_whitelist.c
index 526907d2d8246,cab1b578ca0ef..0000000000000
--- a/drivers/gpu/drm/xe/xe_reg_whitelist.c
+++ b/drivers/gpu/drm/xe/xe_reg_whitelist.c
@@@ -5,6 -5,8 +5,8 @@@
  
  #include "xe_reg_whitelist.h"
  
+ #include <kunit/visibility.h>
+ 
  #include "regs/xe_engine_regs.h"
  #include "regs/xe_gt_regs.h"
  #include "regs/xe_oa_regs.h"
@@@ -41,7 -43,7 +43,7 @@@ static bool match_multi_queue_class(con
  	return xe_gt_supports_multi_queue(gt, hwe->class);
  }
  
- static const struct xe_rtp_table_sr register_whitelist = XE_RTP_TABLE_SR(
+ VISIBLE_IF_KUNIT const struct xe_rtp_table_sr register_whitelist = XE_RTP_TABLE_SR(
  	{ XE_RTP_NAME("WaAllowPMDepthAndInvocationCountAccessFromUMD, 1408556865"),
  	  XE_RTP_RULES(GRAPHICS_VERSION_RANGE(1200, 1210), ENGINE_CLASS(RENDER)),
  	  XE_RTP_ACTIONS(WHITELIST(PS_INVOCATION_COUNT,
@@@ -104,6 -106,7 +106,7 @@@
  				   RING_FORCE_TO_NONPRIV_ACCESS_RW))
  	},
  );
+ EXPORT_SYMBOL_IF_KUNIT(register_whitelist);
  
  static const struct xe_rtp_table_sr oa_whitelist = XE_RTP_TABLE_SR(
  

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

^ permalink raw reply	[flat|nested] 12+ messages in thread

* linux-next: manual merge of the drm-xe tree with the drm-fixes tree
@ 2026-07-03 15:44 Mark Brown
  0 siblings, 0 replies; 12+ messages in thread
From: Mark Brown @ 2026-07-03 15:44 UTC (permalink / raw)
  To: Thomas Hellström, DRM XE List
  Cc: Gustavo Sousa, Linux Kernel Mailing List, Linux Next Mailing List,
	Matt Roper, Violet Monti

[-- Attachment #1: Type: text/plain, Size: 1642 bytes --]

Hi all,

Today's linux-next merge of the drm-xe tree got a conflict in:

  drivers/gpu/drm/xe/xe_tuning.c

between commit:

  e23fafb8594ea ("drm/xe/rtp: Add struct types for RTP tables")

from the drm-fixes tree and commits:

  5ff004fdc7377 ("drm/xe/rtp: Add struct types for RTP tables")
  e9845449e37f5 ("drm/xe/rtp: Ensure gt_was doesn't evaluate rules with engine types")

from the drm-xe tree.

I fixed it up (see below) and can carry the fix as necessary. This
is now fixed as far as linux-next is concerned, but any non trivial
conflicts should be mentioned to your upstream maintainer when your tree
is submitted for merging.  You may also want to consider cooperating
with the maintainer of the conflicting tree to minimise any particularly
complex conflicts.

diff --combined drivers/gpu/drm/xe/xe_tuning.c
index bf3fad9cdbef4,bcec40ca2d35f..0000000000000
--- a/drivers/gpu/drm/xe/xe_tuning.c
+++ b/drivers/gpu/drm/xe/xe_tuning.c
@@@ -20,7 -20,7 +20,7 @@@
  #undef XE_REG_MCR
  #define XE_REG_MCR(...)     XE_REG(__VA_ARGS__, .mcr = 1)
  
- static const struct xe_rtp_table_sr gt_tunings = XE_RTP_TABLE_SR(
+ VISIBLE_IF_KUNIT const struct xe_rtp_table_sr gt_tunings = XE_RTP_TABLE_SR(
  	{ XE_RTP_NAME("Tuning: Blend Fill Caching Optimization Disable"),
  	  XE_RTP_RULES(PLATFORM(DG2)),
  	  XE_RTP_ACTIONS(SET(XEHP_L3SCQREG7, BLEND_FILL_CACHING_OPT_DIS))
@@@ -101,6 -101,7 +101,7 @@@
  				   BANK_HASH_4KB_MODE))
  	},
  );
+ EXPORT_SYMBOL_IF_KUNIT(gt_tunings);
  
  static const struct xe_rtp_table_sr engine_tunings = XE_RTP_TABLE_SR(
  	{ XE_RTP_NAME("Tuning: L3 Hashing Mask"),

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

^ permalink raw reply	[flat|nested] 12+ messages in thread

* linux-next: manual merge of the drm-xe tree with the drm-fixes tree
@ 2026-07-03 16:20 Mark Brown
  0 siblings, 0 replies; 12+ messages in thread
From: Mark Brown @ 2026-07-03 16:20 UTC (permalink / raw)
  To: Thomas Hellström, DRM XE List
  Cc: Gustavo Sousa, Linux Kernel Mailing List, Linux Next Mailing List,
	Matt Roper, Violet Monti

[-- Attachment #1: Type: text/plain, Size: 3821 bytes --]

Hi all,

Today's linux-next merge of the drm-xe tree got a conflict in:

  drivers/gpu/drm/xe/xe_wa.c

between commit:

  e23fafb8594ea ("drm/xe/rtp: Add struct types for RTP tables")

from the drm-fixes tree and commits:

  5ff004fdc7377 ("drm/xe/rtp: Add struct types for RTP tables")
  e9845449e37f5 ("drm/xe/rtp: Ensure gt_was doesn't evaluate rules with engine types")
  e2cfc5bc0c3ff ("drm/xe/rtp: Ensure oob_was does not evaluate engine type rules")
  94e15e89f491e ("drm/xe/rtp: Ensure device_oob_was only evaluates correct rules")
  02b41333f4874 ("drm/xe/xe3p_lpg: Add missing references to workarounds")

from the drm-xe tree.

I fixed it up (see below) and can carry the fix as necessary. This
is now fixed as far as linux-next is concerned, but any non trivial
conflicts should be mentioned to your upstream maintainer when your tree
is submitted for merging.  You may also want to consider cooperating
with the maintainer of the conflicting tree to minimise any particularly
complex conflicts.

diff --combined drivers/gpu/drm/xe/xe_wa.c
index b9d9fe0801aad,139434946f8f8..0000000000000
--- a/drivers/gpu/drm/xe/xe_wa.c
+++ b/drivers/gpu/drm/xe/xe_wa.c
@@@ -130,7 -130,7 +130,7 @@@
  __diag_push();
  __diag_ignore_all("-Woverride-init", "Allow field overrides in table");
  
- static const struct xe_rtp_table_sr gt_was = XE_RTP_TABLE_SR(
+ VISIBLE_IF_KUNIT const struct xe_rtp_table_sr gt_was = XE_RTP_TABLE_SR(
  	/* Workarounds applying over a range of IPs */
  
  	{ XE_RTP_NAME("14011060649"),
@@@ -293,7 -293,7 +293,7 @@@
  	  XE_RTP_ACTIONS(SET(MMIOATSREQLIMIT_GAM_WALK_3D,
  			     DIS_ATS_WRONLY_PG))
  	},
- 	{ XE_RTP_NAME("14026144927, 16029437861"),
+ 	{ XE_RTP_NAME("14026144927, 16029437861, 14026127056"),
  	  XE_RTP_RULES(GRAPHICS_VERSION(3510), GRAPHICS_STEP(A0, B0)),
  	  XE_RTP_ACTIONS(SET(L3SQCREG2, L3_SQ_DISABLE_COAMA_2WAY_COH |
  			     L3_SQ_DISABLE_COAMA))
@@@ -307,6 -307,7 +307,7 @@@
  	  XE_RTP_ACTIONS(SET(GUC_INTR_CHICKEN, DISABLE_SIGNALING_ENGINES))
  	},
  );
+ EXPORT_SYMBOL_IF_KUNIT(gt_was);
  
  static const struct xe_rtp_table_sr engine_was = XE_RTP_TABLE_SR(
  	/* Workarounds applying over a range of IPs */
@@@ -586,12 -587,12 +587,12 @@@
  
  	/* Xe3p_LPG*/
  
- 	{ XE_RTP_NAME("22021149932"),
+ 	{ XE_RTP_NAME("22021149932, 14026290593"),
  	  XE_RTP_RULES(GRAPHICS_VERSION(3510), GRAPHICS_STEP(A0, B0),
  		       FUNC(xe_rtp_match_first_render_or_compute)),
  	  XE_RTP_ACTIONS(SET(LSC_CHICKEN_BIT_0_UDW, SAMPLER_LD_LSC_DISABLE))
  	},
- 	{ XE_RTP_NAME("14025676848"),
+ 	{ XE_RTP_NAME("14025676848, 14026270459"),
  	  XE_RTP_RULES(GRAPHICS_VERSION(3510), GRAPHICS_STEP(A0, B0),
  		       FUNC(xe_rtp_match_first_render_or_compute)),
  	  XE_RTP_ACTIONS(SET(LSC_CHICKEN_BIT_0_UDW, LSCFE_SAME_ADDRESS_ATOMICS_COALESCING_DISABLE))
@@@ -802,10 -803,11 +803,11 @@@ static const struct xe_rtp_entry oob_wa
  
  static_assert(ARRAY_SIZE(oob_was_entries)  == _XE_WA_OOB_COUNT);
  
- static __maybe_unused const struct xe_rtp_table oob_was = {
+ VISIBLE_IF_KUNIT __maybe_unused const struct xe_rtp_table oob_was = {
  	.entries = oob_was_entries,
  	.n_entries = ARRAY_SIZE(oob_was_entries),
  };
+ EXPORT_SYMBOL_IF_KUNIT(oob_was);
  
  static const struct xe_rtp_entry device_oob_was_entries[] = {
  #include <generated/xe_device_wa_oob.c>
@@@ -813,10 -815,11 +815,11 @@@
  
  static_assert(ARRAY_SIZE(device_oob_was_entries) == _XE_DEVICE_WA_OOB_COUNT);
  
- static __maybe_unused const struct xe_rtp_table device_oob_was = {
+ VISIBLE_IF_KUNIT __maybe_unused const struct xe_rtp_table device_oob_was = {
  	.entries = device_oob_was_entries,
  	.n_entries = ARRAY_SIZE(device_oob_was_entries),
  };
+ EXPORT_SYMBOL_IF_KUNIT(device_oob_was);
  
  __diag_pop();
  

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

^ permalink raw reply	[flat|nested] 12+ messages in thread

end of thread, other threads:[~2026-07-03 16:20 UTC | newest]

Thread overview: 12+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2026-07-03 15:44 linux-next: manual merge of the drm-xe tree with the drm-fixes tree Mark Brown
  -- strict thread matches above, loose matches on Subject: below --
2026-07-03 16:20 Mark Brown
2026-07-03 15:44 Mark Brown
2026-07-03 15:41 Mark Brown
2026-07-03 15:41 Mark Brown
2026-01-30 16:05 Mark Brown
2026-01-23 15:34 Mark Brown
2025-09-19 12:49 Mark Brown
2025-09-19 12:45 Mark Brown
2025-09-22 18:28 ` Nathan Chancellor
2025-09-22 19:25   ` Lucas De Marchi
2024-12-06 12:55 Mark Brown

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox